How power-sharing includes and excludes non-dominant communities: Introduction to the special issue

Who and what is represented in a consociation?

At an etymological level, a consociation refers to ‘a society of societies, or a people of peoples’ (O’Leary, 2005a: xxii). It is for this reason that many consociationalists prefer the adage divided places rather than divided societies – in a consociation, the state is always composed of more than one society. From this starting point, power-sharing advocates contend that the transition from war-to-peace has to commence with addressing the substantive concerns of conflict protagonists – or the different peoples in the state – in order to pave the way for peace and democratisation. Such an arrangement is, at a minimum, better than continued intergroup violence and is thus preferable to other institutional solutions. A ‘negotiated consociational settlement’, O’Leary (2005b: 10) argues, ‘is better than a winner-takes-all outcome – especially when taking all implies killing, expelling, or assimilating the losers’. Both Agarin and Stojanović investigate the implications of this starting point and cut right to the core of power-sharing theory by interrogating its foundational assumptions as a theory of democracy.

Timofey Agarin (Agarin, this issue) is interested in the parallels between consociational and liberal democracies both in terms of the manner in which they facilitate participation and in terms of what is being represented in the political process. Focusing on citizens affected by, yet having no guaranteed representation in, power-sharing politics, Agarin tests the limits of institutions designed to put an end to intergroup conflict through representing the diversity of societal identities. His contention is that even if power-sharing settings encourage intergroup cooperation, they do so mainly with the formerly conflicting groups in mind, making non-dominant groups and their identities redundant in the halls of government. While the diversity of identities is recognised as the ‘norm’ of politics in both consociational and liberal democracies, Agarin argues, a focus on the representation of groups deemed ‘significant’ constrains the input of minority and non-dominant groups in both systems.

Nenad Stojanović (Stojanović, this issue) picks up directly on the notion of a ‘people of peoples’ by aligning consociational democracy with the literature on demoi-cracy, which also assumes ‘a Union of peoples who govern together but not as one’ (Nicolaïdis, 2013: 351). While most theorists distinguish between demoi-cracy and consociational democracy, Stojanović identifies some considerable conceptual affinity. His concern is with whether the design of political institutions around competing demoi favours stability over democracy. Ultimately, Stojanović concludes that a (corporate) consociational democracy is ‘a conceptual impossibility’. The demos problem is ‘solved’ in one of two ways in a consociation: either by falling into the ‘ethnocracy trap’ whereby democracy becomes untenable, or by the integration of competing demoi into a single demos, at which point governance is democratic but no longer consociational. A consociation becomes either a more extreme version of ethnic identity politics, as seen in ethnocracies, or it is rendered obsolete if and when ethnic politics yield place to the kind of interest-based politics found in demoicracies and other multinational liberal democracies.

Both articles suggest at the very least, that consociations are resistant to reform because the groups reaping the power-sharing dividend are mindful that their interests are recognised by and vested in the system. According to both authors, as long as power is shared between ethnic groups, consociations appear similar to ethnocracies in their functional modality of ‘government or rule by a particular ethnic group’ (Anderson, 2016), or indeed several ethnic groups. As a result, the members of the political community who do not belong to any one of these segments are factually reduced to second-order partners in political decision-making. While it is possible for a consociation to ‘render itself superfluous’ (Lijphart, 1977), a process Stojanović argues has happened in the early consociational cases of Switzerland and Austria, more often than not it morphs into ethnocratic form, with Bosnia and Herzegovina as the prototypical example. In such cases, consociational institutions, particularly corporate ones, do not become superfluous but instead entrench divisions. As Stojanović demonstrates, direct democracy and other centripetal strategies may be able to contribute to the emergence of a shared identity of peoples as citizens of the same polity. Agarin meanwhile argues that this process is likely to result in the overrepresentation of a newly constituted majority and, as is the case in liberal democracies, its eventual domination over all other non-dominant groups in the citizenry. Agarin and Stojanović, then, in line with the other authors in this special issue, stress the implications of the path-dependent nature of power-sharing institutions, particularly as it relates to how non-dominant groups can participate and gain representation in consociational spaces.

Exclusions in EAI

The second set of articles challenges the assumption that all citizens seek political representation on ethnic terms. Where politics rely on ethno-political bargaining, gender and sexual minorities, micro-minorities, members of groups identifying around non-ethnic or civic labels all find it disproportionally difficult to mobilise around issues that cut across the politically salient group boundaries. Yet, the question of ensuring the political participation of these non-dominant groups is critically important when designing political institutions. In post-conflict settings, the exclusion of such groups can result in the limited acceptance of shared interests, identities and institutions, can prevent conflict resolution by pitting the winners and losers of peace-building against one another, and can curtail the consolidation of democratic principles, all of which threatens to impede the peace-to-democracy transition. This concern is taken up by McCulloch, Byrne, Murtagh and Juon who each, in highlighting particular manifestations of exclusion, also identify strategies for minimising the effects of that exclusion.

Both Allison McCulloch (McCulloch, this issue) and Siobhan Byrne (Byrne, this issue), for instance, consider how women as a constituency have been largely excluded from power-sharing arrangements. While they chart different strategies for minimising the EAI dilemma for women, both authors challenge the view that a gender perspective is necessarily at odds with power-sharing practices. Whereas much feminist scholarship tends to frame power-sharing as gender-blind and exclusionary of women’s interests, McCulloch and Byrne both worry that a blanket rejection of power-sharing risks leaving women without the tools needed to challenge their exclusion. McCulloch suggests that the resources for integrating a gender perspective can be found within contemporary developments in power-sharing theory and practice. In particular, she notes that the widening of power-sharing to new areas can open up new political opportunities for women to participate in the larger power-sharing structure. McCulloch further finds that the distinction between liberal and corporate consociational arrangements remains complicated from a gender perspective, given that many of the typical strategies for including women rely on a kind of ascriptive logic of their own. Instead she argues for institutional strategies that are strong enough to provide the necessary protections for non-dominant groups but which are still liberal enough to avoid both gender and ethnic identity reification. Byrne meanwhile exposes the limited theorisation of power in contemporary power-sharing theory and queries whether critical feminist thinking can help us move past coercive and statist conceptions of power. She finds inspiration in theories of intersectionality and in the proposals brought forward by feminist scholar-practitioners, including notions of responsibility-sharing and sharing-power. Ultimately, by turning to the feminist ethic of care literature, she makes the case for a greater emphasis on the sharing logic in power-sharing, which has largely been ignored by power-sharing scholars. Doing so, Byrne suggests, provides the basis for a ‘new language of both power and sharing’ that takes seriously grassroots resistance, exposes power inequalities, and expands power beyond ethnonational elite interests.

Cera Murtagh (Murtagh, this issue) too takes as her starting point what she sees as a long-standing oversight in power-sharing theory: the endurance of civic parties under a set of institutional rules designed to reward ethnic parties. Disputing the claim that civic parties are theoretically marginal, she presents a complex picture of the fate of such parties in power-sharing systems. As might be anticipated, Murtagh first argues that the political opportunity structure of consociationalism serves as a barrier to the inclusive messaging of civic parties. These barriers can be both formal – corporate rules ensuring ethnic proportionality disadvantage parties seeking to mobilise on a non-ethnic basis – and symbolic – voters are socialised into a system of ethnic representation and see no other alternative than to continue to vote for ethnic parties. Nonetheless, Murtagh also suggests that civic parties have been presented with critical openings, including overrepresentation or balance-of-power roles in governing coalitions. Importantly, and in line with McCulloch, Juon, and Stojanović, she finds that the liberal-corporate distinction matters. Liberal rules enable civic parties to participate on civic terms whereas corporate rules require creative navigation, including sometimes having to act as ethnic representatives in order to gain entry into the system.

Andreas Juon (Juon, this issue) similarly argues that the distinction between liberal and corporate rules has important implications, though he comes at the EAI problem from the perspective of micro-minorities, ethnic groups that make up but a fraction of the overall population. In a large-N study, Juon uncovers a positive outcome for both dominant minorities and micro-minorities under power-sharing rules. Yet, as he suggests, these findings present constitutional designers with a quandary: opt for corporate rules, thereby ensuring strong guarantees for a small number of groups, possibly promoting some groups at the expense of others, or go with liberal rules, which offer weaker provisions but do so for a wider number of groups. While Juon points to a possible remedy of the EAI dilemma, at least for excluded ethnic groups, it is not likely to be popular with dominant parties, who may want to lock in their share of power via corporate rules.

Consociational theory assumes that because ‘ethnic divisions are resilient rather than rapidly biodegradable’ (McGarry and O’Leary, 1995: 338), conflict management is best achieved through the direct political accommodation of the central segments of society. Whilst pragmatic, it leaves open the door to the adoption of corporate over liberal mechanisms for groups’ representation. However, corporate consociations establish a system of representation that reserves place in politics to groups ahead of the political participation of the electorate, in so doing projecting interests upon groups, rather than upon members. At the same time, consociations emphasise the value of, and help to facilitate, free and fair elections. If, as in liberal consociations, no seats or government offices are reserved for specific groups ahead of elections, one might be able to reasonably anticipate that the political salience of major ethnic electoral blocs could disappear or weaken after extended periods of political stability. Should voters be more supportive of candidates who articulate their interests, rather than those who proffer (only) identity group rhetoric, consociational politics could equally advance issues that crosscut cleavages and appeal to non-dominant identity groups.

Inclusions in EAI

The three final articles in this special issue turn the EAI dilemma around to analyse the challenges that exist for dominant minority groups when they enter into power-sharing arrangements with majority groups. That is, despite their formal inclusion, exclusionary practices may well persist for dominant minorities. Dominant minorities, then, experience their own kind of EAI dilemma.

The central argument presented by Eduardo Wassim Aboultaif (Aboultaif, this issue) is that without careful constitutional design, power-sharing is not always sufficient to overcome the tendency towards hegemonic control. First, the balance of power that exists between groups is likely to skew negotiations in favour of the majority; without strong institutions, such as veto rights, the system will struggle to ward off this tendency to control. This has been the case in Lebanon, with the Christian Maronites historically serving as the hegemon even prior to independence (Winslow, 2012: 65). This led to the under-representation of the dominant Muslim minority during the National Pact era, which arguably precipitated the civil war (Salamey, 2013: 56) and which was only partially rectified by the terms of the 1989 Taif Accords. Aboultaif speaks to two aspects of Lijphart’s original conceptualisation of consociationalism. While he confirms that all four institutions need to be adopted as a package in order to achieve stability – a point from which Lijphart later stepped away – Aboultaif disputes Lijphart’s claim that semi-consociations can still promote democratic values. Defined by the concentration of executive powers at the centre and the communal control over the armed forces, semi-consociations allow the communal hegemon in the system to subordinate the interests of all other groups. Yet he sees the glass as half-empty, rather than half-full. Discussing the cases of pre-war Lebanon and post-invasion Iraq, Aboultaif identifies these three mechanisms as the reason for enduring instability and ultimately conflict in these countries.

At the root of his argument is the point that elites must be willing to cooperate with – and not control – one another for power-sharing to work effectively, a point that also emerges in Jon Fraenkel’s analysis. Fraenkel (Fraenkel, this issue) confirms this assumption, demonstrating how elites often enter into power-sharing arrangements for reasons other than a sincere desire to facilitate minority inclusion. As one of the five cases discussed in his article shows, the rules for parliamentary representation in Fiji between 1997 and 2006, which differentiated between communal and ‘cross-voting’ constituencies, was put into place to encourage non-ethnic voting of the multi-ethnic electorate. However, even in light of institutional incentives to de-ethnicise the political party landscape, ethnic Fijian parties put only ‘token’ representatives of the Indian community on their largely Fijian party lists, invariably losing support from Fiji Indian voters and vice versa. Fraenkel observed similar practices in Bosnia, South Africa, and New Caledonia. The Fijian experience highlights how political elites try to circumvent the constraints of the political system in order to avail themselves of what is, in effect, the consolidation of ethnic politics.

Also stemming from the willingness of elites to cooperate is the question of what inclusion brings for those minorities who are part of a power-sharing coalition. Fraenkel demonstrates how the EAI dilemma can also exist for included minorities. That is, instead of the inclusion of some groups resulting in the exclusion of others, he finds that inclusion and exclusion can become simultaneous processes for included groups. When minority representatives join power-sharing coalitions, they risk, first, being perceived as sell-outs by rival parties outside of the coalition, jeopardising their support levels at the next election, and second, being marginalised within cabinet and not being able to pass legislation in support of their own interests. Reading Aboultaif and Fraenkel in conjunction, there is a powerful argument for revisiting power-sharing’s assumption about the positive role political elites play in facilitating the transition from peace to democracy. Political elites’ positioning at the centre of power, their intimate knowledge of the political process, and their prime position to make political capital out of the peace dividend make them unlikely agents of change in the transformation from identity- to issue-based politics.

While Aboultaif and Fraenkel highlight the challenges of inclusion for dominant minorities, Charis Psaltis, Huseyin Cakal, Işık Kuşçu and Neophytos Loizides (Psaltis et al., this issue) push the conceptualisation of the EAI dilemma in new directions with a case study of the exclusion of IDPs from the Cyprus peace process. In their article, they challenge the assumption that ethnic parties speak for all members of the ethnic group and highlight how the differential experiences of IDPs and non-IDPs during the conflict informs their perspectives on the contours of a future settlement. Psaltis et al. reach two important conclusions. First, drawing on new survey data collected shortly after the failed Crans-Montana talks in 2017, they find that, against the extant literature on refugees and conflict resolution, IDP status in Cyprus is associated with higher levels of support for the peace process. Second, their survey data also demonstrate the importance of getting the institutional proposals right. Even when support for general principles of power-sharing is forthcoming, as in a bicommunal, bizonal federation, this does not automatically extend to support for specific institutional recommendations, as reservations on the proposed presidential arrangement indicate.

Consociational theory posits that the war-to-peace transition will eventually be followed by a peace-to-democracy transition. As this happens, the relevance of sectarian identities should dissipate and yield place to cooperative interest-based politics. The central thrust of consociationalism is thus to strengthen the authority of the state, while integrating all those who have input into and benefit from the development of a shared polity into the decision-making process. Yet, as Murtagh, Aboultaif and Fraenkel suggest, political cooperation in many deeply divided places often contests, rather than upholds, the ability of policy instruments to reach across the divide and shape meaningful narratives of social and political change without references to the group identities to be represented in consociational politics. If, as Psaltis et al. argue, resolving protracted conflicts is done in the name of ‘the people’, the representatives of all affected groups should be involved in the political process. To this end, consociational institutions should seek to demonstrate that group identity markers, rather than being permanently fixed, can contribute to and reflect the growth of shared interests, which can and should speak to all groups – and individuals – affected by the peace-to-democracy transition.